Research Work | Mysite

Influence of uncertainties on the seismic life cycle cost assessment of the RC building

Seismic life cycle cost assessment (S-LCCA) incorporates lifetime seismic-hazard induced repair costs in terms of present-day values and provides insights on the need for an immediate seismic upgrade, lifetime cost and benefits associated with the structure. The state-of-the-art methodology of S-LCCA incorporates three essential components namely seismic fragility assessment, seismic hazard occurrence model and repair cost models. However, as a consequence of different sources of uncertainty prevailing in each component, estimation of S-LCCA may vary substantially. This study aims to quantify and propagate uncertainties stemming from such sources and assess their influence on S-LCCA estimation.

Machine learning algorithms for estimation of RC beam-column model parameters and uncertainty quantification

In order to simulate the nonlinear response lumped plasticity beam-column element model has been widely adopted. The numerical modeling parameters for an element are estimated using the trilinear backbone curve. At present, a widely used approach for estimating modeling parameters utilizes linear regression-based semi-empirical equations. Model uncertainty is described by the constant standard deviation. As significant heterogeneity prevails in the column properties, the assumption of the constant standard deviation throughout the sample space can lead to inaccurate estimation of the uncertainty in this approach. This study aims to help researchers, stakeholders, and decision-makers to select appropriate machine learning algorithms to predict model parameters that are simple to implement yet aids in reasonable quantification of response uncertainty under seismic shaking.